Object detection and tracking for Personal Protective Equipment

This paper focuses on real-time detection and tracking techniques for Personal Protective Equipment (PPE) in workplace scenarios. Based on the critical characteristics of efficiency, the potential solution addresses the identification and tracking of PPE for safety and anti-injury purposes by leveraging advanced underlying technologies. Therefore, this paper presents a real-world applicability, efficiency, and reliability in PPE monitoring using YOLOv8 (You Only Look Once version 8). The overall approach of the system includes data cleaning, model learning, and deployment. The efficiency of the proposed model is evaluated by performance indicators precision, recall, etc. The findings thus support the applicability of YOLOv8 in real-scenario settings with reliable PPE identification and monitoring. The accuracy achieved with the proposed model is ${9 4 \%}$.